'Children's cafeterias' in Japan hit record 12,601 sites, survey reveals The number of children's cafeterias that provide free or low-cost meals mainly to children in need in Japan rose by more than 1,700 from the previous fiscal year, according to a survey by a nonprofit organization. The number of children's cafeterias, which provide free or low-cost meals mainly to children in need in Japan, has reached a record 12,601 this fiscal year, according to a survey by a nonprofit organization. The total rose by more than 1,700 from the previous fiscal year, said the survey released Thursday by Musubie, a Tokyo-based nonprofit supporting kodomo shokudō programs nationwide. The nonprofit organization said the expansion reflected efforts by central and local governments to create comfortable spaces for children. We aim to create an environment that makes it easier to start and sustain kodomo shokudō programs, Musubie head Rie Mishima said at a news conference.

Japan's bear-related casualties hit record on escalating attacks Bear bells are displayed for sale at a souvenir shop at Shirakawa-go, a popular tourist spot and one of Japan's UNESCO World Heritage sites, in the village of Shirakawa, Gifu Prefecture, on Nov. 15. A record 230 were killed or injured by bears in Japan since April, putting more pressure on the government to intervene as the animals push deeper into areas where people live. Thirteen have died and 217 were injured as a result of bear attacks in the eight months through end-November, according to data released Friday by the environment ministry. The total already exceeds the previous record of 219 for the fiscal year through March 2024. Roughly two-thirds of casualties occurred in the sparsely-populated northern Tohoku region.

The total number of US Customs and Border Protection device searches jumped by 17 percent over the 2024 fiscal year, but more invasive forensic searches remain relatively rare. Over the Past year, United States Customs and Border Protection staff searched more phones and electronic devices at the border than ever before, according to new statistics published by the government agency. Phone searches jumped around 17 percent during the past 12 months--with a marked increase over the past six months. Newly published CBP figures show that for the full fiscal year of 2025--running from October 2024 to the end of September 2025--border agents conducted around 55,424 searches of electronic devices. This is up from around the 47,000 searches that were completed during the government's 2024 fiscal year.

Transcending human cognitive limitations represents a critical frontier in LLM training. Proprietary agentic systems like DeepResearch have demonstrated superhuman capabilities on extremely complex information-seeking benchmarks such as BrowseComp, a feat previously unattainable. We posit that their success hinges on a sophisticated reasoning pattern absent in open-source models: the ability to systematically reduce extreme uncertainty when navigating vast information landscapes. Based on this insight, we introduce WebSailor, a complete post-training methodology designed to instill this crucial capability. Our approach involves generating novel, high-uncertainty tasks through structured sampling and information obfuscation, RFT cold start, and an efficient agentic RL training algorithm, Duplicating Sampling Policy Optimization (DUPO). With this integrated pipeline, WebSailor significantly outperforms all open-source agents in complex information-seeking tasks, matching proprietary agents' performance and closing the capability gap.

The US's largest companies have spent 2025 locked in a competition to spend more money than one another, lavishing 155bn on the development of artificial intelligence, more than the US government has spent on education, training, employment and social services in the 2025 fiscal year so far. Based on the most recent financial disclosures of Silicon Valley's biggest players, the race is about to accelerate to hundreds of billions in a single year. Over the past two weeks, Meta, Microsoft, Amazon, and Alphabet, Google's parent, have shared their quarterly public financial reports. Each disclosed that their year-to-date capital expenditure, a figure that refers to the money companies spend to acquire or upgrade tangible assets, already totals tens of billions. Capex, as the term is abbreviated, is a proxy for technology companies' spending on AI because the technology requires gargantuan investments in physical infrastructure, namely data centers, which require large amounts of power, water and expensive semiconductor chips.

Microsoft is now the second company ever to surpass 4 trillion in market valuation, following artificial intelligence giant Nvidia. Microsoft, which is traded under the ticker "MSFT", is continuing to surge and as of noon in New York City (16:00 GMT) on Thursday, it is up 4.6 percent from the market open. The technology behemoth said it will spend 30bn in capital spending for the first quarter of the current fiscal year to meet soaring artificial intelligence (AI) demand. Microsoft also reported booming sales in its Azure cloud computing business on Wednesday. "It is in the process of becoming more of a cloud infrastructure business and a leader in enterprise AI, doing so very profitably and cash generatively despite the heavy AI capital expenditures," said Gerrit Smit, lead portfolio manager, Stonehage Fleming Global Best Ideas Equity Fund. Redmond, Washington-headquartered Microsoft first cracked the 1 trillion mark in April 2019.

Microsoft recently released the Microsoft Digital Defense Report 2024, this year's edition of the company's annual cybersecurity report. In the 114-page document, Microsoft reveals -- among other things -- just how much cyber threats have grown over the past year. Cybercriminals have gained access to better resources, including the incorporation of AI tools to bolster their arsenal. They're now better equipped to create fake images, videos, and audio recordings to trick people, to flood job applications with AI-created "perfect" résumés to physically access companies, and much more. But hackers can also use your use of AI services to attack you.

We investigate whether an LLM can successfully perform financial statement analysis in a way similar to a professional human analyst. We provide standardized and anonymous financial statements to GPT4 and instruct the model to analyze them to determine the direction of future earnings. Even without any narrative or industry-specific information, the LLM outperforms financial analysts in its ability to predict earnings changes. The LLM exhibits a relative advantage over human analysts in situations when the analysts tend to struggle. Furthermore, we find that the prediction accuracy of the LLM is on par with the performance of a narrowly trained state-of-the-art ML model. LLM prediction does not stem from its training memory. Instead, we find that the LLM generates useful narrative insights about a company's future performance. Lastly, our trading strategies based on GPT's predictions yield a higher Sharpe ratio and alphas than strategies based on other models. Taken together, our results suggest that LLMs may take a central role in decision-making.

The financial domain frequently deals with large numbers of long documents that are essential for daily operations. Significant effort is put towards automating financial data analysis. However, a persistent challenge, not limited to the finance domain, is the scarcity of datasets that accurately reflect real-world tasks for model evaluation. Existing datasets are often constrained by size, context, or relevance to practical applications. Moreover, LLMs are currently trained on trillions of tokens of text, limiting access to novel data or documents that models have not encountered during training for unbiased evaluation. We propose SEC-QA, a continuous dataset generation framework with two key features: 1) the semi-automatic generation of Question-Answer (QA) pairs spanning multiple long context financial documents, which better represent real-world financial scenarios; 2) the ability to continually refresh the dataset using the most recent public document collections, not yet ingested by LLMs. Our experiments show that current retrieval augmented generation methods systematically fail to answer these challenging multi-document questions. In response, we introduce a QA system based on program-of-thought that improves the ability to perform complex information retrieval and quantitative reasoning pipelines, thereby increasing QA accuracy.

Samsung has failed to recover from the sharp decline in profit it experienced in 2022. In its latest earnings report, the Korean company has reported KRW 258.94 trillion ( 194 billion) in annual revenue and KRW 6.57 trillion ( 4.9 billion) in operating profit for the fiscal year of 2023. Those are markedly smaller numbers than the previous fiscal year's, especially the latter's -- Samsung posted an operating profit of KRW 43.38 trillion ( 35 billion) for 2022, which was already 6.9 billion smaller than the year before due to the weak demand for its chips and smartphones. According to The Wall Street Journal, these numbers represent Samsung's weakest earnings in over a decade. The company says its memory business showed signs of recovery, but not enough to stop it from incurring KRW 2.18 trillion ( 1.63 billion) in operating losses for the fourth quarter of 2023.